Key Characteristics of Cardiovascular Toxicants

According to the World Health Organization, cardiovascular disease (CVD) is the leading cause of death worldwide, taking $1.7\text{\hspace{0.17em}million}$ lives annually (WHO 2017). Four of five of those deaths are due to myocardial infarction or stroke. Certain environmental pollutants, such as fine particulate matter [PM$\le 2.5\mathrm{\mu }\mathrm{m}$ in aerodynamic diameter (${\mathrm{PM}}_{2.5}$)] (Brook et al. 2010, 2016), arsenic (States et al. 2009) and tobacco smoke (Gallucci et al. 2020), are well known to be associated with CVD, but other environmental contaminants, as well as natural toxins, viruses, and other agents, may also be cardiovascular (CV) toxicants.

A systematic approach to identifying chemical hazards was recently developed for carcinogens (Smith et al. 2016), endocrine-disrupting chemicals (La Merrill et al. 2020), and reproductive toxicants (Arzuaga et al. 2019; Luderer et al. 2019) based on the established properties of chemicals known to cause cancer, endocrine disruption, and reproductive toxicity, respectively. These properties, called key characteristics (KCs), have quickly proved useful for the systematic evaluation of the literature on mechanisms by which chemicals induce these toxic effects (Guyton et al. 2018a, 2018b). The KCs are now widely used by various authoritative bodies and regulatory agencies and form the basis for the evaluation of mechanistic data at the International Agency for Research on Cancer (IARC 2019; Samet et al. 2020). Scientists in the pharmaceutical industry have also recognized that the KCs are likely to be useful in the design of a comprehensive set of tests to evaluate the potential hazards of novel drug candidates (Fielden et al. 2018; Smith et al. 2020).

Our goal was to develop a consensus on the KCs of chemical and nonchemical agents known to cause CV toxicity and to provide a comprehensive list of tests to be used to evaluate chemicals and other environmental pollutants for CV toxicity. Given that many pharmaceutical drugs have adverse effects on the CV system and because those mechanisms are generally better understood than those of environmental pollutants, we included data from pharmaceuticals in the development of the KCs of CV toxicants. As outlined in Figure 1, we believe there are multiple ways in which these KCs of CV toxicants could be used to enhance current approaches in the clinic and in pharmaceutical development, environmental research, and hazard assessment.

Experts from various fields related to CV toxicity and chemical regulation convened and identified 12 KCs of CV toxicants using current scientific evidence, such as the earlier work of Laverty et al. (2011), expert knowledge, known examples of CV toxicants, and extensive debate. We acknowledge that these will likely evolve with new scientific discoveries. In considering the differences between acute and chronic effects, as well as between high- and low-dose effects, we concluded that these KCs cover temporal and dose-dependent cardiotoxic mechanisms. We did, however, restrict our task to adult CVD, excluding possible teratogenic effects of environmental pollutants on the developing CV system.

We also identified representative biomarkers, assays, and end points that are most useful for testing each KC using experimental in vitro/ex vivo studies and in vivo animal models, as well as clinical or epidemiological findings in humans (Table 1). Further, we identified classic examples of CV toxicants for each KC (Table 1) and illustrated how some CV toxicants exhibit multiple KCs, whereas other toxicants may exhibit only one (Tables 2 and 3). We have divided the KCs into those primarily affecting cardiac tissue (numbered 1–4), vascular tissues (5–7), and those which could affect both the heart and vasculature (8–12).

Table 1, which is based on the expertise of the authors, summarizes established assays and other methods to evaluate the 12 different KCs in vitro, in animal models, and in humans; see also a summary of evaluation methods by Berridge et al.(2013). From these summaries, it is evident that a) although some assays used in vitro (e.g., cytotoxicity assays, quantitative polymerase chain reaction) are high-throughput and, therefore, suited to serve as screening for multiple chemicals, such high-throughput assays are not available for most KCs (Villeneuve et al. 2019); b) there are usually fewer end points that can be measured in humans, and those are most often not high-throughput; and c) in some cases, the output of the assays and end points used in vitro and in animals are not easily translated into the output of the biomarkers used in humans.

Regarding KC1 (cardiac excitability), robust in vitro functional assays and end points (e.g., half maximal inhibitory concentration values) are available to determine altered ${\mathrm{Na}}^{+}$ or ${\mathrm{K}}^{+}$ channel function (Mathie et al. 2021; Sigg et al. 2010); in nonrodent animal models and humans, the fast sodium and hERG channels are linked directly to ventricular depolarization (e.g., QRS interval) and repolarization (e.g., QT interval), respectively (Baldrick 2021; Edwards and Louch 2017; Strauss et al. 2021). The evaluation of intracellular ${\mathrm{Ca}}^{2+}$ concentrations in humans is not currently feasible in clinical practice (Bruton et al. 2020). For KC2 (contractility and relaxation), ultrasound of the heart is a very commonly used technique both in animal models and in humans (Lindsey et al. 2018; Wang et al. 2018). Cardiomyocyte injury and death for KC3 can be evaluated in vitro and in animals in several ways. Cardiac troponins are a very useful and rather high-throughput biomarker in humans regarding myocardial infarction (Taggart et al. 2021), but a good assay to evaluate apoptosis is needed in the clinical setting (Mohamad Kamal et al. 2020). Echocardiography, cardiac magnetic resonance imaging, and myocardial perfusion scan can, however, provide evaluation of overall cardiac function and areas of myocardial injury (Makavos et al. 2021; Sivapackiam et al. 2020; Sreenivasan et al. 2021). Regarding the proliferation of valve stroma for KC4, a high-throughput assay is available for in vitro use (Reid et al. 2013), and echocardiographic imaging in humans is routinely performed to evaluate valvular heart disease (Jain et al. 2021). Endothelial and vascular function for KC5 can be assessed with isolated blood vessels obtained from animals [medium-throughput (Knox et al. 2019)] and human donors [low-throughput (Virdis and Taddei 2016)]. It could be studied in humans in vivo by flow-mediated vasodilation (Tremblay and Pyke 2018), but this technique is only reliable in younger subjects (Lind 2006). Mitochondrial function for KC8 can be measured well in vitro (Koklesova et al. 2021), but is difficult to directly evaluate in humans (Pelletier-Galarneau et al. 2021). There are several techniques (e.g., heart rate variability) to measure ANS activity (Nolte et al. 2017) and understand sympathetic and parasympathetic involvement for KC9 that are relatively easy to apply in humans (Cygankiewicz and Zareba 2013) but more challenging in animal models. Measurements of biomarkers related to hemostasis (KC6) (Adelborg et al. 2021; Lind et al. 2011), dyslipidemia (KC7) (Lind 2019; van Wijk et al. 2009), oxidative stress (KC10) (Kumar et al. 2014b; Tejchman et al. 2021), inflammation (KC11) (Friedman and Shorey 2019; Kumar et al. 2014a; Libby 2002), and hormone signaling (KC12) (Penell et al. 2021; Svobodová and Cajthaml 2010) are readily achievable in both experimental models and humans.

We therefore conclude that there is an unmet need for a systematic improvement of the nonclinical assays, biomarkers, and physical tests used to evaluate all of the KCs. There is also a need to standardize nonclinical tests to assure data quality and reproducibility, as well as their value for translation to human investigations. Hence, the systematic and comprehensive identification of the KCs and the available end points presented herein will help to prioritize the development of improved methods to evaluate potential CV toxicants both experimentally and in humans. Ideally, qualified biomarkers could be used to advance public health by assisting regulatory decision-making (FDA 2019).

Figure 2 illustrates how the KCs may contribute to the pathogenesis of acute and chronic injury to the heart (Figure 2A) and blood vessels (Figure 2B). Note that multiple KCs may contribute at different locations in the CV system to produce short- or long-term injury and eventually disease. Below and in Tables 2 and 3 we detail how the KCs can be used to generate a holistic picture of how environmental pollutants and drugs that are established CV toxicants can cause CV toxicity. We also describe how the KCs can contribute to understanding the effects of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). These examples further illustrate how evidence for each KC can be organized and evaluated using the published literature.

Regulatory agencies consider a broad range of health end points when determining if a drug or an exogenous chemical poses a hazard. Given the importance of CVD as a major heath burden on society, it is critical to identify potential environmental CVD hazards and reduce exposure to them. Like the KCs for other organ systems, the 12 KCs described here will help these agencies better evaluate hazards and risks to human health by facilitating the systematic assessment of the mechanistic data (Figure 1). In the area of clinical practice, the KCs can help to target improvements in assays, biomarkers, and physiological tests used for risk assessment and differential diagnoses. For toxicologists, the KCs provide a potential framework to facilitate a holistic approach to studies of the potential effects of both pharmaceutical drugs and environmental chemicals on CV toxicity through in vitro screening, in vivo characterization, and human data. Further, the identification of KCs and knowledge of the methods to evaluate them will inform the development of high-throughput assays and in silico screens that could be used to expedite acquisition of information regarding potential CV toxicity (Blanchette et al. 2019; Burnett et al. 2021; Sirenko et al. 2017). The KC framework also enables study of the CV effects of mixtures comprising chemicals that exhibit different KCs, as was recently described for studies of the carcinogenic effects of mixtures (Rider et al. 2021).

Development of the 12 KCs described herein benefited substantially from experience with pharmaceutical drugs, by taking advantage of knowledge collected over decades of drug development and intensive research on adverse effects of pharmaceuticals. We have applied this list of KCs to both pharmaceutical agents and environmental pollutants to illustrate the utility of this approach for both classes of chemicals. However, not all possible mechanisms whereby a chemical could cause CV toxicity may be covered by these 12 KCs, and it is likely that as knowledge in the field of CV toxicity advances, more KCs could be added in the future. For example, emerging evidence suggests that exposure to PM in air pollution impacts the human gut microbiome, resulting in an altered intestinal redox lipidome that may contribute to CV and gastrointestinal diseases (Feng et al. 2020). However, further elucidation is needed to determine whether this represents a characteristic of CV toxicants, as well as whether it is sufficiently distinct from other KCs, such as KC7.

The examples of SARS-CoV-2 and air pollution show that the KCs also could be used for nonchemical agents and mixtures, respectively, and not just individual chemicals. Other such examples include food items, such as foods high in saturated fat (Saluja et al. 2021) and sweetened beverages that have been linked to CVD (Kim et al. 2020), radiation therapy that could give rise to a great number of myocardial pathologies ranging from fibrosis to ischemic disease (Burke et al. 2020), and the well-known endocarditis induced by Streptococcal infections (Chamat-Hedemand et al. 2020). Because food items are a mixture of naturally occurring and synthetic chemicals, the KC approach could be used to investigate which components of the food may possess CV toxicant properties.

It also should be emphasized that a KC is akin to an umbrella under which many different detailed pathophysiological events take place. These events and their underlying molecular mechanisms are understood to varying degrees. For some KCs, such as effects on the ${\mathrm{Na}}^{+}{/\mathrm{K}}^{+}$ ion channels and the valvular stroma, the critical mechanisms are known at the molecular and cellular levels, but for KCs such as inflammation and oxidative stress, much knowledge on basic pathophysiology is yet to be uncovered in detail. The KC approach represents a systematic way to not only define the pathophysiological events involved in CV toxicity but also to identify and address gaps in the mechanistic understanding of how exogenous chemicals could harm the heart or vasculature.

Some environmental pollutants, such as PCBs, arsenic, and dioxins, have been linked to both CVD (Li et al. 2015; Lind et al. 2019) and cancer (Lauby-Secretan et al. 2016). The identification of KCs for the major organ systems, or health conditions, could facilitate identification of important mechanisms whereby chemicals could affect multiple different organ systems. Thus, if a KC were found to be linked to impairments in several organ systems or pathological processes, such as cancer development, the need to develop high-throughput testing assays for this KC would be of higher priority. In a similar manner, although our focus was on CV toxicity in adults in this commentary, KCs relevant to the developing organ systems, including the CV system, and associated health conditions could be developed in the future.

In conclusion, the identification of KCs for CV toxicants facilitates, in an objective and systematic fashion, the further understanding of CV effects of exogenous chemicals, chemical mixtures, and other agents, both in drug development as well as for environmental pollutants. When coupled to the development of high-throughput assays, new candidate drugs, novel marketplace chemicals, as well as existing agents to which the population is exposed, could be screened for specific unwanted characteristics. Thus, the systematic use of KCs may serve as a powerful tool to prevent and mitigate CV toxicity.

This project was supported by contracts 17-E0023 and 18-E0034 from the Office of Environmental Health Hazard Assessment of the California Environmental Protection Agency (EPA) and the Research Translation Core of the University of California (UC), Berkeley National Institute of Environmental Health Sciences (NIEHS) Superfund Research Program (SRP) under National Institutes of Health (NIH) grant P42ES004705 (M.T.S.); Project 4 of the UC Davis NIEHS SRP under NIH grant P42 ES004699 (A.V.G. and N.C.); the Texas A&M SRP under NIH grant P42 ES027704, Texas A&M NIEHS Center grant P30 ES029067, and a cooperative agreement with the U.S. EPA, STAR RD83580201 (W.A.C.); NIH grant 5R01HL139472 (N.G.P.); NIH grant R01ES023470 (C.Z.), and NIEHS grant R01 ES029395 and R01ES032806 (J.A.A.). The opinions are those of the authors and do not reflect official views of their respective agencies.

The views expressed in this publication are those of the authors and do not necessarily represent the decisions, policy or views of their respective institutions. Reference to commercial products or services does not constitute endorsement or recommendation for use. This manuscript has been reviewed by the Center for Public Health and Environmental Assessment, U.S. EPA, and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the agency nor does mention of trade names or commercial products constitute endorsement or recommendation for use.

H.M.V. is an employee of Amgen, Inc. All other authors declare they have no actual or potential competing financial interests.